Skip to content
MCP server

Toolc

Compiler that merges many MCP servers and REST APIs into one lean tool endpoint

Runs locallyStatus and history
Pricing
Open source
Platforms
CLI, self-hosted, API
Runs
Locally
Version
0.1.1
Listing updated
today

About Toolc

Toolc is an open-source MCP server for ai models and llm tooling. It runs locally on your machine and is started by your AI client.

Connecting several MCP servers to one agent floods the model's context with tool definitions and makes it worse at picking the right one. Toolc attacks that by treating the whole set of downstream servers as input to a compiler: you point it at MCP servers, an OpenAPI spec or even a documentation page, and it serves a single reshaped tool surface in return.

The passes are individually switchable. They rewrite tool descriptions so sibling tools are told apart, fold near-duplicate tools such as search, find and get variants into facade tools with deterministic routing, and can hide the catalogue behind two meta-tools, search_tools and call_tool, for maximum context savings. At serve time oversized results are summarised while keeping identifiers, and every call is logged with its arguments and results.

The README reports a benchmark of 131 tools across five live servers where the compiled setup kept 96.7% success at about 6.4 times lower cost per completed task. The LLM-assisted passes need an Anthropic API key; without one the server mirrors the original tool surface. A hosted version of the same engine also exists.

Install

Run in a terminal

claude mcp add toolc --env ANTHROPIC_API_KEY='<ANTHROPIC_API_KEY>' -- npx -y @toolc/cli

Add to claude_desktop_config.json

{
  "mcpServers": {
    "toolc": {
      "command": "npx",
      "args": [
        "-y",
        "@toolc/cli"
      ],
      "env": {
        "ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>"
      }
    }
  }
}

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "toolc": {
      "command": "npx",
      "args": [
        "-y",
        "@toolc/cli"
      ],
      "env": {
        "ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>"
      }
    }
  }
}

Add to .vscode/mcp.json

{
  "servers": {
    "toolc": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@toolc/cli"
      ],
      "env": {
        "ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>"
      }
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "toolc": {
      "command": "npx",
      "args": [
        "-y",
        "@toolc/cli"
      ],
      "env": {
        "ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>"
      }
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "toolc": {
      "command": "npx",
      "args": [
        "-y",
        "@toolc/cli"
      ],
      "env": {
        "ANTHROPIC_API_KEY": "<ANTHROPIC_API_KEY>"
      }
    }
  }
}

Configuration

  • ANTHROPIC_API_KEYEnables the LLM-assisted compile passes (rewrite, consolidate). Without it, toolc serves a mirror of the original tool surface.

Tools this server provides (2)

Taken from the project documentation; the running server may differ.

search_tools
Meta-tool that searches the compressed tool catalogue.
call_tool
Meta-tool that calls a tool found through search.

What it does

Description rewriting
Rewrites tool descriptions for agents and disambiguates similar sibling tools.
Tool consolidation
Merges families of near-duplicate tools into single facade tools with deterministic routing.
Search-based selection
Compresses the catalogue behind search_tools and call_tool to save context.
Result compaction
Summarises oversized results at serve time while keeping identifiers and sources.
REST composition
Adds REST APIs from an OpenAPI spec or a spec drafted from documentation.
Call logging
Logs every call with arguments and results and clusters recurring errors into suggested fixes.

Good for

  • Combine a dozen MCP servers behind one endpoint for an agent
  • Cut tool-definition token cost for a production agent
  • Expose a REST API with an OpenAPI spec as MCP tools
  • Audit what an agent actually called and why it failed

Pricing

Open source

Engine is Apache-2.0; a managed hosted service also exists, prices not published

Prices change. Check toolc.dev before you buy.

Toolc FAQ

Does it need an API key?

An ANTHROPIC_API_KEY enables the rewrite and consolidate passes; without it toolc serves a mirror of the original tool surface.

Is it open source?

The engine is Apache-2.0 licensed, and a managed hosted service runs the same engine.

How is it run locally?

Clone the repo, install with pnpm, write a config, then run compile and serve over stdio; the npm package is @toolc/cli.

Can it front REST APIs?

Yes, a downstream can be an OpenAPI spec as well as an MCP server.

Is Toolc free?

Yes. Toolc is open source, so you can use and self-host it for free.

How do I add Toolc to Claude Code?

Run this command in your terminal: claude mcp add toolc --env ANTHROPIC_API_KEY='<ANTHROPIC_API_KEY>' -- npx -y @toolc/cli

What are the best alternatives to Toolc?

The closest alternatives to Toolc are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

Alternatives to Toolc

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools